Classifications

• • G—PHYSICS
  • G06—COMPUTING; CALCULATING OR COUNTING
  • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
  • G06V20/00—Scenes; Scene-specific elements
  • G06V20/50—Context or environment of the image
  • G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle

Definitions

• the present invention relates generally to driving aids for motor vehicles.
• This process comprises steps:
• the driver may also be dazzled by elements that are in the field of the headlights and reflect the light from high beams in the direction of the driver.
• the calculation of this level of glare is generally performed taking into account the luminance of the light source seen by a camera equipping the motor vehicle, and the solid angle according to which the light source is seen by this camera.
• the luminance is for example estimated according to the average of the color codes associated with the pixels which, on the image acquired by the camera, correspond to the light source.
• the solid angle is obtained according to the surface occupied by the light source on the acquired image and according to the model of the camera, which takes into account its intrinsic and extrinsic parameters through a geometric calibration.
• the determination of the glare level according to this method is effective as long as the distance between the camera and the light source remains reduced. However, when this distance increases, the calculation of the level of glare is less accurate, so that the control of the obscuration of glasses glasses is less reliable.
• the Applicant has in fact understood that, as long as the light source occupies an important place on the acquired image, the calculation of the solid angle can be carried out with precision. On the other hand, as soon as this space becomes too small (for example when the light source is represented by only one pixel on the acquired image), this calculation must no longer be considered reliable.
• the present invention proposes a method for improving the estimation of the level of glare when the distance between the light source and the motor vehicle is high.
• the invention proposes not to improve the algorithm for estimating the distance given the shape of the light source on the acquired image, but on the contrary to take advantage of the presence on the vehicle of another type of detector to determine with greater precision the distance between the vehicle and the light source. It proposes in particular to take advantage of this detector when the distance between the light source and the image sensor is greater than 20 meters (in the case where the image sensor comprises a wide-angle lens) or 100 meters (in the case where the image sensor has a longer focal length lens).
• the estimation step it is intended to estimate a surface occupied by said light source on said image, and, at the calculation step, said glare level is also calculated according to said surface;
• said area is estimated taking into account the results of measurements made by said detector;
• the calculation step it is intended to determine a first estimate of the solid angle according to which the light source is seen by the image sensor, as a function of said distance, and then to determine a first estimate of the level dazzle according to the first estimate of said solid angle;
• the model comprises intrinsic parameters of the image sensor (which vary according to the type of sensor chosen), such as, inter alia, the vergence, the distortion introduced by the lens of the image sensor, and the transformation between the image sensor system; coordinates whose image sensor is at the origin and the coordinates of the pixels in the image;
• the model also comprises extrinsic parameters (which vary according to the position of the image sensor in space) such as the translation vector and the rotation matrix which define the position and to the orientation of the image. image sensor in the vehicle;
• said level of glare is calculated as a function of the luminance of the light source as it appears in the image
• said distance is estimated by the detector by means of a RADAR or LIDAR detection technique.
• the invention also proposes a motor vehicle comprising:
• a detector adapted to measure a distance separating the motor vehicle from a source of light illuminating a driver of the motor vehicle
• a computer adapted to communicate with the image sensor and the detector, for implementing a determination method as mentioned above.
• FIG. 1 is a schematic perspective view of a motor vehicle according to the invention.
• FIGS. 2 and 3 are representations of images that can be acquired by an image sensor fitted to the motor vehicle of FIG. 1.
• a motor vehicle 10 which is here in the form of a four-wheeled car 1 1.
• a motor vehicle comprising two or three wheels, or more wheels.
• this motor vehicle 10 comprises a chassis which notably supports a powertrain 12 (namely a motor and means for transmitting torque from the engine to the driving wheels), body elements and cabin elements.
• a powertrain 12 namely a motor and means for transmitting torque from the engine to the driving wheels
• the motor vehicle 10 also comprises an electronic control unit (or ECU, of the English “Electronic Control Unit”), here called calculator 14.
• ECU electronice control unit
• This calculator 14 comprises a processor and a storage unit, for example a rewritable non-volatile memory or a hard disk.
• the storage unit notably records computer programs comprising instructions whose execution by the processor enables the computer to implement the method described below.
• the computer 14 is connected to different equipment of the motor vehicle 10.
• the motor vehicle 10 comprises at least one image sensor 15 and a distance detector 1 6. It further comprises a communication means 18 adapted to communicate with a microcontroller fitted to a pair of glasses 17 carried by the driver of the vehicle.
• the image sensor is here formed by a camera 15 which is oriented forward, so that it can acquire images of a portion of the road at the front of the vehicle. It may be a so-called wide-angle camera (that is, a camera that has a wider viewing angle than the natural perception of the human eye) or a camera with a focal length long (ie a camera that has a lens whose focal length is larger than that of a wide-angle lens).
• a camera 15 which is oriented forward, so that it can acquire images of a portion of the road at the front of the vehicle. It may be a so-called wide-angle camera (that is, a camera that has a wider viewing angle than the natural perception of the human eye) or a camera with a focal length long (ie a camera that has a lens whose focal length is larger than that of a wide-angle lens).
• This camera 15 is here represented as being fixed in the front bumper of the vehicle. Alternatively, especially if it is of type "long focal length", the camera could be located at the rear of the windshield 13 of the vehicle. Alternatively, this camera could be mounted on the vehicle removably. It could thus be a so-called embedded camera. Alternatively, it could be the camera on the mobile phone of the driver.
• the camera 15 must be able to acquire images of a portion of the road at the front of the camera. vehicle and to communicate information to the calculator 14 of the vehicle.
• the camera can be of any type, the fact remains that it will be necessary to perform a photometric calibration to find the correspondence between the luminance level (cd / m 2 ) of the light source and the values of the pixels in the image to obtain comparable results regardless of the camera used. It will thus be necessary to calibrate it taking into account the model of this camera, which considers the intrinsic and extrinsic parameters of this camera (notably the position of this camera in relation to the eyes of the driver).
• the distance detector 1 6 may also be in various forms and may be placed in various locations of the vehicle. However, it will be positioned so as to detect obstacles located at the front of the vehicle, in the field of view of the camera 15.
• This distance detector 1 6 will preferably be designed to emit waves toward the front of the vehicle, to detect the waves reflected by the obstacles at the front of the vehicle, and to determine the distance between the vehicle and these obstacles, given the time taken by the waves to return to the distance detector 1 6.
• this distance detector 1 6 can operate according to one or the other of the following technologies:
• - LIDAR (acronym for "light detection and ranging”, ie “detection and localization by light”); it may thus be for example a laser remote sensor;
• the present invention relates to estimating a glare level of the driver to reduce the discomfort felt by the driver. Otherwise formulated, the present invention is intended to provide a level of glare that can be used to control an anti-glare means.
• the windshield 13 is equipped with a film adapted to darken at least partially, under the control of the computer 14.
• the computer 14 could thus control this obscuration as a function of the level of glare detected.
• the driver wears a pair of glasses 17 whose lenses are adapted to darken, under the control of a microcontroller fitted to this pair of glasses.
• the computer will thus be able to control the obscuration of the spectacle lenses according to the level of glare detected, by controlling the microcontroller accordingly via a communication means 18 fitted to the vehicle and via an antenna equipping the pair of spectacles.
• the motor vehicle 10 is shown in Figure 1 as rolling on a road with different infrastructure, including a road sign 2 located on the edge.
• the invention then relates to a method implemented by the computer 14 of the motor vehicle 10 to determine the level of glare caused by the light sources dazzling the motor vehicle.
• the sources of light considered may then have different shapes. It can thus notably be:
• the light source dazzling the driver is formed either by the road sign 2 located on the edge of the road, or by the front lights 4, 5 of a vehicle traveling in the opposite direction.
• the computer 14 is then designed to implement a method of determining the level of glare generated by this road sign 2 or by these front lights 4, 5.
• the road sign 2 and the front lights 4, 5 will be called "light sources 2, 4, 5".
• the determination process has four main steps, including:
• the solid angle ⁇ is calculated as a function of at least the distance L separating the camera 15 from the light source 2, 4, 5.
• the distance L is estimated as a function of the measurement results made by the detector 16. In this way, the calculation of the solid angle is very accurate, which reduces the risk of error in determining the glare level of the driver.
• the aforementioned four steps are performed regularly by the computer 14, at reduced time intervals, so as to obtain continuously a good estimate of the glare level of the driver.
• the camera acquires an image 1 of a portion of the road located at the front of the motor vehicle 10.
• Examples of images 1 acquired are for example shown in Figures 2 and 3. It is observed in these images 1 that, since the vehicle is driving at night, the traffic sign 2 or the front lights 4, 5 stand out clearly from the rest of the image (which is here represented by simplification by a black background).
• the computer 12 receives from the camera 15 this image 1, which it processes so as to distinguish the zone (s) 3 corresponding to light sources 2, 4, 5 from the rest of the background.
• This second step can be implemented in different ways. It can thus, for example, consist in applying a filter to the image, which operates by thresholding. In this example, all the pixels of the image having a color code greater than a threshold are considered potentially dazzling and are distinguished from the background of the image. These pixels are then gathered together (s) of related pixels, each set defining a zone 3 in the sense of the invention.
• the image 1 has only one zone 3.
• it comprises two corresponding to the two headlights 4, 5 of the vehicle traveling in the opposite direction.
• the calculator 14 calculates an estimate of the solid angle ⁇ according to which the light source 2, 4, 5 is seen by the camera 15.
• L is the distance separating the motor vehicle from the light source, and is expressed in meters.
• the calculator 14 here calculates three estimates of this solid angle ⁇ .
• the first estimate of this solid angle ⁇ is calculated in a manner already known from the state of the art, as a function of the place occupied by the light source 2, 4, 5 on the image 1 acquired and as a function of the model of the camera 15. Since this method is already known, it will not be described here in more detail.
• the distance L is directly obtained by the distance detector 1 6.
• the surface S ' corresponds to the actual surface of the traffic sign 2 (as opposed to the surface S of the zone 3 of the image 1). It can be obtained in various ways.
• the computer may have in memory a database associating with each type of light source a surface S '(for example 1, 1 m 2 for a triangular traffic sign, 0.15 m 2 for a front light of car, 0.1 m 2 for a car front light, .
• the computer 14 may for example take into account one and / or the other of the following parameters:
• the frequency of the illumination intensity of the light source (the latter being different for a car headlight and a luminaire along the road).
• the first solution (which consists in deducing the surface S 'from the surface S and from the distance L) can be chosen as long as the distance L is less than a threshold.
• the second solution (which consists of using a database) can be chosen when the distance L exceeds this threshold.
• the third estimate of the solid angle ⁇ will be calculated according to the results obtained via the first and the second estimate of this solid angle ⁇ .
• the distance L will have been estimated, on the one hand, according to the data appearing on the acquired image 1, and on the other, according to the data received from the distance detector 1 6.
• This average will be weighted to take into account the confidence that is given to the measurement made by the camera 15 and the one that is given to the measurement made by the distance detector 1 6.
• the computer 14 determines the level of glare B caused by the reflection of the light coming from the front lights of the motor vehicle 10 on the road sign 2.
• the calculator operates here the following calculation:
• the luminance L can for example be obtained by averaging the luminances of the pixels of the zone 3 of the image 1, that is to say by considering the color codes of these pixels.
• the luminance L b can be obtained in the same way, considering the remaining part of the image 1. However, to obtain comparable results regardless of the camera used, it is necessary to calibrate this camera.
• This calibration process can for example be performed by positioning the camera in a dark room inside which is a calibrated light source. Thus, by illuminating the camera with a known luminous intensity (and luminance), it is possible to associate this intensity (or luminance) with the measurement made by the camera.
• the angle ⁇ is obtained given two phenomena.
• the first phenomenon is that the human eye is more dazzled when it observes directly a source of light than when it observes a point offset from this source of light.
• the angle ⁇ will be obtained as a function of the distance separating, on the one hand, the center of the image 1, and on the other hand, the center of the zone 3 considered from the image 1.
• the second phenomenon is that the glare felt at the camera 15 is not the same as that felt by the driver, given the fact that the camera is not positioned at the eyes of the driver. This angle ⁇ will then be obtained depending also on the position of the camera on the vehicle.
• the level of glare B obtained it can be used in various ways.
• the computer 14 determines a darkening setpoint for the glasses of the pair of spectacles 17 worn by the driver.
• the computer 14 is then programmed to communicate this instruction to the microcontroller equipping the pair of glasses 17, so that it controls the darkening of the glasses of the pairs of glasses, which prevents the driver from being dazzled.
• the computer 14 in the case where the glare is due to the reflection of the light of the headlights of the vehicle considered on the signaling panel 2, it would be possible for the computer 14 to communicate to the headlights a setpoint for reducing their luminous intensity (or , if the headlights are movable, an instruction to change the direction of illumination of the headlights).
• the present invention is not limited to the embodiment described and shown, but the art can apply any variant within his mind.
• the glare level could be estimated based solely on two parameters, namely the luminance of the signaling panel on the image and the distance separating the light source from the camera.
• the luminances L, L b used in the calculation of the glare level could be, not read on the image acquired by the camera, but obtained by means of light intensity detection cells fitted to the vehicle.

Landscapes

• Engineering & Computer Science (AREA)
• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Multimedia (AREA)
• Theoretical Computer Science (AREA)
• Traffic Control Systems (AREA)

Applications Claiming Priority (2)

Publications (2)

Family

ID=57121333

Family Applications (1)

Country Status (3)

Families Citing this family (3)

Family Cites Families (4)

• 2016
  • 2016-07-19 FR FR1656875A patent/FR3054356B1/fr active Active
• 2017
  • 2017-07-19 EP EP17748676.8A patent/EP3488380B1/de active Active
  • 2017-07-19 WO PCT/EP2017/068250 patent/WO2018015446A1/fr unknown

Also Published As

Similar Documents

Legal Events